THE SOCIOLOGICAL STUDY of CRIMINALITY Thorsten Sellin

Journal of Criminal Law and Criminology

Volume 41 | Issue 4 Article 3

1951 The oS ciological Study of Criminality Thorsten Sellin

Follow this and additional works at: https://scholarlycommons.law.northwestern.edu/jclc Part of the Criminal Law Commons, Criminology Commons, and the Criminology and Criminal Justice Commons

Recommended Citation Thorsten Sellin, The ocS iological Study of Criminality, 41 J. Crim. L. & Criminology 406 (1950-1951)

This Article is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for inclusion in Journal of Criminal Law and Criminology by an authorized editor of Northwestern University School of Law Scholarly Commons. THE SOCIOLOGICAL STUDY OF CRIMINALITY Thorsten Sellin

The following is the general report for Sociology as presented at the Second International Congress of Criminology in Paris, September 12, 1950. The author is professor and chairman of the Department of Sociology in the University of Penn- sylvania. Since 1929 he has been editor of the ANNALS OF THE AMERICAN ACADEMY OF POLITICAL AND SOCIAL SCIENCE. Since January, 1950, he has been absent on leave in Berne, Switzerland, while serving as Secretaiy General of the International Penal and Penitentiary Commission. During many years he was consultant in criminal statistics to the Bureau of the Census. From 1927 to 1937 he was Associate Editor of this JOURNAL in charge of the Book Review Section. He continues as a member of our Advisory Editorial Council. Professor Sellin's latest contribution to our pages is in Volume 40, Number 6 (March-April 1950) under the title: "The Uniform Criminal Statistics Act."-EDITOR.

I. The program of this Congress calls for exclusive attention to the problems of the etiology of crime and methods of research in the un- covering of causal relationships in criminal conduct. It has set as one of its aims a kind of inventory of etiological research in the various disciplines concerned, an inventory leading to a statement of what may today be considered as established knowledge rather than speculations and guesses. The making of such an assessment, however, presents great difficulties in some disciplines, because they lack adequate media of communication, while in others such media are excellent and permit the researcher to keep in constant touch with research developments in his field. The social sciences generally are at a great disadvantage in this respect. The result is that the individual sociologist interested in criminal etiology-usually not exclusively so-finds himself greatly frustrated in his desire to keep abreast of research done in his field throughout the world. He is often compelled to rely on what is done in his own country or published in his own language. This encourages an insular viewpoint and deprives him of the intellectual stimulus which is gained by a wider acquaintance with the work of his colleagues in other countries. It is not implied that this situation is peculiar to sociology. It is more or less common in all disciplines so far as criminological study is concerned. We invite examination of the two encyclo- pedias of criminology published in recent years in Italy and the United States for the most clear evidence of that fact. Coming back to sociology, it was hoped that the assessment of research findings on an international scale could be made on the basis of an adequate number of national reports. Unfortunately this hope has been found to be vain. In spite of the existence of institutes of criminology or criminological societies in many countries, large or small, 406 1950] STUDY 0F CRIMINOLOGY only three national reports have been presented, one from Turkey, prepared by Dr. Nurullah Kunter, Professor of Criminal Law, Uni- versity of Istanbul; one from Venezuela, by Dr. Jos6 Rafael Mendoza, Titular Professor of Criminal Law, Central University of Venezuela; and one from the United States by Dr. Marshall B. Clinard, Professor of Sociology, University of Wisconsin." A report submitted from France is not a survey of the type conceived of by the program committee but a report on a single research project recently completed, dealing with the offense of "abuse of confidence." In fact, the report from Turkey, aside from excellent observations on methodology and the quality of sources of data there available-especially official statistical sources- is devoted largely to an analysis of the results of a recent study of 6,386 prisoners convicted of homicide. The Venezuelan report is a most interesting monograph on criminality in Venezuela with a general discussion of its extent and causes, based on official criminal statistics. The authors of both these reports have been handicapped by the fact that sociological research on the etiology of crime in their countries appears to be rare. Nevertheless, they have presented very interesting communications, which we hope will be printed in the proceedings of this Congress. Since under the circumstances, only casual references will be made to these communications in this report, it is hoped that both Dr. Kunter and Dr. Mendoza will supply these defects by the presentation of their views during the discussion in the section. A report which would be limited almost exclusively to sociological research in the United States, could hardly be considered a general report of the type expected by the participants in the Congress. Since the rapporteur, for many reasons, has been unable in the time at his disposal even to attempt to supply the deficiencies due to the absence of national reports from a dozen or more countries where much important sociological research has been done, he has chosen to avail himself of the suggestion of the Secretariat, to deal with the general topic of the report in any convenient way. In so doing he will not fail to refer to the few reports submitted whenever possible. He offers in advance an apology to other rapporteurs, should he at times trespass on their preserves.

II.

When we examine the history of research on criminal conduct we observe very quickly that this history is nothing but one aspect of the more general history of the study of the determinants of human behavior

1. Dr. Clinard's report will be published later in this JOURNAL.-EDITOR. THORSTEN 8ELLIN[ [Vol. 41

in general. The desire for scientifically valid explanations of why human beings behave in this way or another has not been confined to any one specialized science. It has stimulated enquiry by research workers in such diverse disciplines as sociology, psychology, psychiatry, ethnology, economics, neurology, physiology, anatomy, genetics, anthropology (both physical and cultural), biochemistry, biophysics and some others. Each of these sciences has been and is cultivated by persons who have received a basic professional education in the social, biological, or physical sciences. This training has given the scientist a fundamental equipment of concepts, attitudes and frames of reference for his work as well as techniques of investigation which permit him to develop greater competency in a still more specialized area of research. Itis characteristic of all etiological science that it seeks to discover constant relationships among the phenomena investigated, arriving there- by at the formulation of generalizations or laws that permit prediction and hence open the possibility for control. Research into the etiology of criminal conduct shares this aim, for without knowledge of the processes or the concatenating conditions that give rise to such conduct, prediction and control, i.e. prevention or consciously directed modification, are impossible. Another characteristic of scientific research is that each science operates on the basis of fundamental assumptions, useful and necessary for the investigation of the phenomena with which it is concerned. "Criminal anthropology," says Di Tullio, for instance, "can, as a biological discipline, but follow, more and more rigorously, those very methods that the biological sciences in general follow in the study of the normal, abnormal, or sick person . . . and has to apply and utilize (the methods) of those other biological disciplines with which it has the strongest relations. ' 2 This means then that when a person trained in the biological sciences approaches the study of the etiology of crime he is compelled to employ the scientific knowledge he has mastered and the techniques of investigation proper to his field. The same is true of the sociologist, the psychologist, the medical scientist and the rest of the students of criminal conduct. The basic assumptions and the techniques of investigation utilized by one of these sciences may not serve the purpose or be acceptable to the others, although they may all focus their attention on the same problem, and all employ the inductive- deductive logic common to all sciences. It is gratifying that the problem of criminal conduct has engaged the interest of workers in so many different disciplines. This reflects the

2. BENIGNO Di TULLIO, ANTROPOLOGIA CRIMINALE. 511 pp. Roma: Luigi Pozzi, 1940, p. 6. 1950] ,STUDY OF CBIMINOLOGY

importance of the social problem of crime and the strength of the desire to discover effective means of solving it or at least of understanding it. This diversity of research effort, however, presents certain draw- backs. One of its results is the prevalence of numerous, often seemingly unrelated and sometimes conflicting theories and opinions on the causes of crime. They are the consequence of the division of scientific labor and of the necessary differences in the starting points for research in the various disciplines. A concomitant result is the development of a scientific terminology within each discipline, words sometimes becoming charged with special meanings which the worker in another discipline finds difficulty in understanding. This is not strange, since even within a single discipline there is often a lack of precision in the use of scientific terms or the use of the same term to denote concepts which are not identical. With the above all too brief observations in mind we may draw the following conclusions: 1. Etiological research on criminal conduct must, in the nature of things, be carried out, within each discipline on the basis of the general assumptions, premises, frames of reference, etc., and by means of the special techniques proper to that discipline and hence learned by its votaries. Such specialization is the characteristic of modern science and is necessary for the growth of knowledge. It follows that useful dis- coveries by a research worker are not likely to be made except in the field of knowledge he has mastered. 2. There is no simple royal road to knowledge. While there are still occasional persons who maintain that crime can be ascribed to a single cause, they must be regarded as at the rear of modern thinking. Each discipline concerned with the explanation of human behavior may be able to contribute knowledge that is valuable now -or in the future. The debate on the superiority of this or that discipline should be closed. The scientific value of research depends on the adequacy of theory and the knowledge and skill of the researcher, no matter in what discipline he operates. The practical value of his findings is another matter. That depends on whether or not a particular society is ready to translate the findings of research into social action, i.e. whether or not the findings are socially, rather than scientifically, acceptable, and thus permit social experimentation. 3. Scientifically valid findings of research in a given discipline is of the greatest value to the research worker in another field who, from his own point of view, is concerned with the same problems. Such findings may influence him in orienting and planning his own researches and 410 THOIRSTEN SELLIN [Vol. 41. in interpreting their results. This expansion of his intellectual horizon is no modest task. Unless it is based on serious effort and stimulated by intellectual cooperation it often becomes superficial. We observe, rather as a rule than as the exception, the uncritical use of borrowed data because the borrower is not in a position to assess properly the value of such data because he lacks adequate knowledge about the field from which they were drawn. 4. The cross-fertilization of ideas just mentioned can be rendered more fruitful by: (a) Team-work, in which research workers from different fields join in the investigation of a specific problem, each bringing his specialized knowledge to bear upon it and all members of the team agreeing on a division of labor and on basic principles; (b) meetings of representatives of different disciplines for the discussion of common problems, a device illustrated by this Congress of Criminology; and (c) an international abstract service which would be a medium for reporting findings of etiological research on criminal conduct no matter in what discipline they originate. These are all devices for improving the communication of ideas. They would help to break down the isolationism which has so often characterized scientific research in criminal etiology. It is this isolationism coupled with a failure to understand the history of science and the consequences of the division of scientific labor, which has caused the frequent misunderstandings and criticisms leveled at each other by investigators in different fields. 5. Since criminal conduct is being studied by investigators in so many disciplines, armed with such different basic assumptions, knowledge and techniques, it is idle to spend time on discussing whether or not there is a science of criminology. Whatever scientific knowledge we possess about the etiology of criminal conduct consists of what sociologists, psychologists, psychiatrists, etc. have gathered when they have, often due to a sporadic, temporary or incidental concern with the problem, focused their attention on criminal conduct. Such heterogeneous data do not permit of synthesis into a theoretical whole at present and debate is not likely to produce it, nor ingenious definitions. If we call this aggregate of ideas "criminology" let us frankly recognize that this term has only a practical utility. The same holds true for the taxonomies or "systems" of "criminology" which now and then find their way into print. From a scientific point of view they have no 1950] STUDY OF CRIMINOLOGY

special value, even though they may be of practicalvalue in setting up teaching programs or in planning a congress, for instance. The advancement of our knowledge does not depend on "criminologists," a title about as meaningful as that of doctor of science or doctor of philosophy, but on sociologists, psychiatrists, psychologists, etc.,- "criminologists" by courtesy and, in one sense at least, kings without a country. The advancement of knowledge does depend on the degree to which these categories of scientists can formulate and plan more carefully the research projects they undertake; perfect their techniques of investigation and of analysis of results; and, evolve meaningful bodies of theory. If one etiological science of criminology should ever be possible it will emerge in due time.

III. An examination of crime causation research in various countries, in terms of its history and present status, reveals certain facts of interest to the sociologist. In some countries most of such research seems to be concentrated in what might be called the biological sciences, while in others social science research has been equally prominent or has even dominated. The explanation that might most readily be seized upon would be that all behavior sciences have not reached the same degree of cultivation or advancement in all countries. The historical priority of the biological disciplines is evident, and it is equally evident that in some countries sociology is still an undeveloped discipline and has not emerged from its original introspective stage nor has received full standing as an academic subject. This could be assumed to account for the con- centration of research which makes it appear as if now the biological orientation and now the sociological orientation characterized causation study in this or that nation. The amount of causation research within a given discipline in a country obviously is in part related to the research manpower resources within that discipline and in part to the awareness of the problem, i.e. the extent of the desire or need for explaining criminal conduct. Mr. Clinard suggests in his report that in this connection the existence of high crime rates stimulate enquiry. He assumes that such rates make a people keenly aware of criminality and eager to understand its causes and that nations with low crime rates are less aware of it and hence less concerned. Awareness of a problem is undoubtedly necessary to stimulate research, but awareness is a relative concept. A people's awareness of a problem comes when it exceeds a certain limit of tolerance. What one nation will tolerate and remain unaware of depends on its level of THOPSTEN SELLIN [ Vol. 41 civilization, its social organization, its particular culture. A country which in comparison with other countries statistically shows a low crime rate may consider that rate or rates as intolerable in terms of its culture. If the term high crime rates, employed by Mr. Clinard, is understood in that sense there is some justification for holding that they are a stimulus to research. There have been suggestions made that the dominance of research in one rather than in another discipline in a given country depends on the peculiar nature of criminality in that country. It has been claimed that in countries with low crime rates that reflect cultural homogeneity and effective social control the lawbreaker tends to be an abnormal individual, while in countries with pronounced cultural heterogeneity and high crime rates the normal individual plays a larger role in serious criminality. Hence, biologically oriented research would tend to be popular in the former and sociologically oriented research in the latter type of culture. We doubt that this is a correct conclusion and suspect that the popularity of the one or the other orientation of research is due to deeper culturally conditioned attitudes toward behavior determinants in general: attitudes linked to differences in social organization, political and economic ideolo- gies, etc., all factors that have tended to retard sociological and foster biological research in general or vice versa. The history of criminological research within each discipline shows clearly its dependence on the general changes within that discipline in the theoretical formulation of problems and in techniques of investigation. If we limit ourselves to the last century and a half we note that in the biological field the phrenological theories of Gall, the theories of degeneracy of Morel, the concept of moral insanity, the evolutionary theory of Darwin, the concept of morphological types, Mendel's theory of heredity, etc. have put their imprint on researches on criminal conduct too, while in the social sciences various theories of economic and cultural determinism have played a similar role. Awareness of the problem in the biological and psychological disciplines has been stimulated by the necessity for finding better ways of dealing with the mentally defective, the psychotic or psychopathic offender, the very definitions of which have changed as these sciences have developed. In the social sciences, the problems growing out of the industrialization and urbanization of society, the migration of population groups, etc. have stimulated not only general research in those disciplines but also studies on criminal conduct. Each new technique of investigation designed for general purposes in a given discipline has been employed in criminological researches 1950] STUDY OF CRIMINOLOGY as well and has often given such researches a distinctive character. We need but mention the rise of statistical analysis, the development of psychological testing devices, anthropometry, Freudian analytical techniques, and social attitude tests as illustrations.

IV. The sociologist is interested in knowing how human society is organized and how it functions. He notes that human beings form groups or associations and that these groups develop agencies or institutions through which they can achieve certain ends. He observes as incidental to this process the rise of struggles or conflicts, coordination and mal- adjustment, and wonders what principles or laws there may be that govern these myriad events not only in the large but in their infinite variations and details. When he considers one of the attendant problems of social life, crime, the sociologist observes that this is non-conforming conduct castigated by the state, one of the institutions mentioned, and in varying degree and extent by other social groups and he wonders whether or not the explanation of the volume, specific characteristics and modalities of such conduct, whether common to a given social group or as individual activity by some of its members, may be found in the static conditions or the dynamic processes of social life. This is why he is claimed to be concerned with the investigation of social environmental influences on conduct. As a result of this orientation it is natural for the sociologist to study the incidence of criminality or criminals in specific social groups in rela- tion to the composition or character and the conditions or transforma- tions in the life of such groups. He wants to know how the type, composition, cultural origin of the family is reflected in the conduct of its members; how the type of neighborhood affects the conduct of those who live in it; how the level of education, the occupational character of the group, its class or caste status, etc. produce or inhibit criminality among its members. He is interested in the criminogenic effect of social crises like war and economic depression, the conflict of social classes and the breakdown of social control, to mention but a few illustrations. Studies of the type mentioned, while based on mass data, requiring for their successful prosecution not only adequate indexes of criminality but adequate data on the various social or group characteristics mentioned, have obviously been based on the assumption that criminality is conduct by individuals and that the conclusions of the studies would THOR TEN SELLIN [VOL 41

have some value in understanding individual criminal conduct. The sociologist has not been unaware of the fact that he has in fact ignored the specific individual in such studies and that he has indeed given but certain gross generalizations which may have had value for the general understanding of social processes but have had little value in understanding the conduct of any specific individual offender. This has led him to develop techniques of investigation characteristic of many other behavior sciences. The psychiatrist and the psychologist, for instance, has always employed a clinical technique-the study of the individual case-for that is by the nature of things the only technique possible for disciplines that have grown out of medicine. The sociologist has in more recent years begun to work in the same manner. In engaging in the careful study of individual offenders, however, he has been guided by his sociological orientation, although he has undoubtedly been in- fluenced by developments in other behavior sciences. In approaching the individual case the sociologist's basic assumptions can be stated somewhat like this. An individual becomes a member of society by a process of learning. This learning process begins as soon as he, at birth, makes contacts with others, who have already achieved such membership in varying degree and extent and who are the carriers of the ideas, customs, beliefs, attitudes, etc., which they have acquired in the course of their life experiences as members of social groups. Among the influences which are brought to bear on the new social recruit are concepts of morality, ideas of right and wrong conduct, in general or in specific situations, which are accepted by his "teachers." As the individual grows up and enlarges his contacts his moral concepts are fortified or modified and new ones added, depending on the nature of his experiences. In some types of societies, these moral ideas tend to be consistent and pervasive, while in others they exhibit diversity and even mutual disharmonies and conflict. In dealing with the individual offender, then, the sociologist is not concerned with his constitutional type or his intelligence, his endocrine glands or his Oedipus complex but with his socialization process, his character as a member of social groups, his attitudes and his social experiences in general. There is where he seeks the explanation of criminal conduct. It is easy to understand therefore why the sociologist assumes that conduct, the very existence of which is dependent on socially defined norms, cannot be inherited in any biological sense. The sociologist's preoccupation with his own point of view is not to be regarded as any stranger than that which characterizes any other behavior science. He has frequently been found to adopt skeptical and 1950] STUDY OF CRIMINOLOGY sometimes a superior attitude toward the findings of other behavior sciences, but this too is not a singular characteristic but has its counter- part in all disciplines, for it is a human failing to regard one's own work or discipline as being of leading importance. The sociologist is perfectly willing to admit that biological and psychological factors that enter into the make-up of man influence his receptivity to social influences and his adaptation of the environments that become a part of his personality, when transformed into ideas and paterns of ideas. He is eager to know how the socialization process is affected in the individual case by whatever biological or psychological characteristics workers in those sciences may demonstrate the existence of in that type of individual. However, he is competent to do research only in his own field of knowledge, on the basis of what he considers sociological data and by the use of sociological concepts and techniques of investigation. If he has been critical of some of the findings in other behavior sciences, he has generally taken his stand on a methodological level, for scientific logic offers a common ground on which all the sciences meet. On that ground he has also been as vulnerable as his fellow-workers in other disciplines. Sociological research has not taken the same forms in all countries. Theoretical approaches also show interesting differences. It is quite likely that a certain uniformity that has tended to characterize con- tinental studies in this connection in the past is bound up with the existence in Europe of good. criminal statistics. In the United States, which notoriously lack such statistics of any real value, the sociologist has been compelled to rely on other sources of data. This necessity has led to studies which have proved of considerable importance in shaping new conceptions of causation and, unless we are incorrect, in stimulating some research in some other countries, especially in Great Britain, the Netherlands, the Scandinavian countries, Belgium and Germany, during the last decade and a half. Furthermore, sociology has, because of its popularity as an academic discipline, engaged several thousand scholars in the United States and among them many whose interest has guided them into the study of criminal conduct. It may be noted that of the dozen or more textbooks or manuals of criminology in the United States only one has been written by a non-sociologist and that advanced training in criminology is practically limited to the departments of sociology in the postgraduate schools which prepare for higher degrees. This fact more than any other has probably been responsible for the opinion which seems to be held by some European writers that all studies of criminals and their conduct in the United States are made by sociologists THORSTEN SELLIN [Vol. 4 and that the sociological orientation is not only dominant but exclusively so. This is, as a matter of fact, erroneous to some degree. The bibliog- raphy of psychological, psychoanalytic and psychiatric writings by experts in those fields is voluminous and a number of noteworthy works have appeared, especially since the war. Since criminologists in the United States have only one journal, a bi-monthly, which also is devoted to criminal law and police science, they publish most of their works in the professional journals of the different disciplines. These journals are not always accessible to foreign scholars, who are more likely to consult, therefore, the manuals of criminology written by sociologists. Hence the impression previously mentioned is formed. This impression has had another effect too. Some foreign scholars have voiced the opinion that American sociological theories of criminality have risen out of peculiar conditions in the American culture; that American criminals are different from European ones; and that American findings of research are therefore not applicable to European or Hispanic American (Dr. Mendoza) conditions. We believe that such opinions are erroneous. They have generally been voiced by psychiatrists, are based on their more limited experience with chiefly abnormal offenders and rest on somewhat different conceptions of the meaning of the terms crime and criminal. It is undoubtedly true that criminogenic factors in homo- geneous cultures appear to be different from those in heterogeneous cultures, but the differences must be in degree and not in kind. The reports of Messrs. Kunter and Mendoza give ample data which fit into the conceptual schemes of modern sociological criminology and, we suggest, that the development of criminality during and immediately following the war in Europe clearly indicates that American sociological theories of criminality can be used elsewhere and in all probability have universal application. If they did not they should be changed, for theories that cannot be so formulated that they can be universally employed in research are invalid. It would therefore be highly desirable that they be tested in other countries. American sociologists, says Mr. Clinard: . . believe that there is a preponderance of evidence to indicate that crime is a product of definitions of situations acquired in life experience . . . The origin of crime must be sought in definitions which are present in the culture in the form of competing value systems or culture conflict. These competing value systems arise out of disorganization in social institutions and community situations . . . Sociologists . . . are primarily interested in the conflicting values present in our culture, in the extent of a criminal's membership in groups with deviant values, the role the person plays in such deviant groups, his conception of himself, arising out of such group participation, the conduct norms in the neighborhood from which he comes, the extent of his mobility and association with other deviant 1950] STUDY OF CBIMINOLOGY norms, his attitudes toward law and society, and his degree of criminal association . . . According to the view of most sociologists, participation in deviant norms, particularly through the tutelage of others, is the basic situation out of which most crimes and delinquency arise. Supporting evidence may be found in numerous studies ... This approach to crime has not only been supported by a considerable, although still insufficient number of studies of individual criminals and delinquents, but other support has come from the field of social psychology and from the comparative study of cultures. Both Dr. Kunter and Dr. Mendoza present data that can be inter- preted in these terms. Dr. Kunter discusses the close relationship between urbanization and property offenses, the effect of internal migration on the increase of crime, the role in homicide played by the strong sentiment of personal and family honor in village communities, injured by real or fancied insults, and by the custom of abducting brides, which lead to vendettas. Dr. Mendoza offers evidence to show how the social status of women produces a comparative low female crime rate, how common superstitions, secular and religious popular festivals, sport events and play customs produce social frictions leading to violence; how alcohol habits are related to crime and how, generally speaking, crime tends to take violent rather than' fraudulent forms in his country. Mr. Clinard points out that while sociologists in the United States are in general agreement "about the primary importance . . . of acquired attitudes derived from social experience" and "of culture conflict in the explanation of crime" they are not in complete agreement on the extent to which "the differential response patterns of individuals must also be taken into account." Some would exclude psychogenic personality traits, others stress the need for considering them in seeking to account for the conduct of the individual offender. Trends seem to favor the latter view. V. A problem which has been widely discussed in recent years by American sociologists is the need for a sociological definition of crime and criminal, rather than a legal one. The argument has been presented that what the sociologist is primarily concerned with in connection with crime is the fact that it is a violation of a conduct norm. He notes, however, that there are many similar conduct norms in force in different social groups and some of them have as great and sometimes even greater validity as rules of conduct in these groups than have the norms embodied in the criminal law. He rebels at the necessity of confining his research within a frame of reference as arbitrary as that of the law, and feels the need of all scientists to define for himself the basic data THOISTEN SELLIN [Vol. 41 in research, on the assumption that only by so doing may he hope to arrive at universally valid generalizations. We shall not give the details of this argument here nor of the criticisms which it has met. Its effectiveness has, however, been reflected in recent researches in the United States in which a definite attempt has been made to expand the study of legally defined criminal conduct to include conduct which while criminal is generally dealt with by ad- ministrative authorities or civil law courts and hence does not result in convictions in criminal courts. While this is a modest step in the direction of redefining our field of research and reformulating the basic concepts therein, it is a sign of a fundamental discontent among sociological criminologists, who are looking toward the future. The latest sociological textbook on criminology published in the United States, under the influence of this trend, deals not with crime alone, as tradi- tionally defined in law, but with drug addiction, alcoholism, vagrancy, begging and prostitution, which in part do not fall within the criminal 3 law. VI. No discussion of the sociologist's studies on criminal conduct would be complete without some reference to the sources of his data. This must not be confused with his methods or techniques of investigation, which involve the manner in which he extracts data from the sources and manipulates, organizes, and utilizes them. The sociologist must, of course, utilize available criminal statistics, the records of agencies and institutions that deal with offenses and offenders, autobiographical statements of offenders (personal life history documents). These are recorded sources of data. Unrecorded sources are the minds of offenders or of those who have come into contact with them or with offenses. The exploitation of these latter sources is bound up with interviewing techniques. Sources of data which the sociologist needs in order to verify the accuracy of information drawn from the above sources must also be consulted. There is no need to enumerate them. Nor is there need to more than mention the fact that the nature of the problem investigated will determine what sources he will need to use for securing information permitting him to answer the question or solve the problem

3. "A broader interpretation of crime," says Mr. Clinard, "will undoubtedly dispose of generalizations to the effect that crime is usually a product of such factors as poverty, low intelligence, innate viciousness, emotional difficulties, low education, and broken homes. In place of these we shall have to develop a theory of crime which includes all violations of conduct norms and illegal acts committed by persons of the lower as well as the upper socio-economic groups . . . This effort to reformulate the sociological definition of crime constitutes one of the chief contributions of American sociologists to the development of a science of criminology." 1950] STUDY OF GIMINOLOGY

which prompted his search for information. Such sources are limitless. They may yield information on economic or political conditions, population characteristics, etc., etc. It is important to recognize that the content and quality of certain vitally important sources of recorded data circumscribe the sociologists' researches. For instance, generally speaking, the recorded data tend to become less and less detailed the milder the offense. The most complete data are likely to concern serious offenses or offenders. This means also that sources of information are much richer for long term prisoners than for short term ones. One cannot sufficiently stress the fact that most of the studies of offenders have been made on institutionalized longterm prisoners, who are mostly recidivists, and represent adequate samples of only certain categories of offenders; they are in varying degree, depending on the character of their offenses, the sediment deposited by a long selective process. One source of data must be given a special mention, because it has for over a hundred years been perhaps the most exploited of all sources of information about offenses and offenders, namely criminal statistics, for it is in this source that the criminologist has sought the data that would make it possible to construct accurate measures of the incidence or trend of this or that form of criminality either in geographic areas or in population groups within those areas. This is not the proper place to discuss the value and limitations of criminal statistics for the purposes indicated. We wish simply to note the fact that during the last hundred years during which the behavior sciences have made considerable advances, the items of information about offenses and offenders tabulated in criminal statistics have undergone relatively little modification. This cannot be blamed entirely on the official services which collect and publish these statistics. Much of the blame rests on us, whether sociologists, psychiatrists, or psychologists, for although we are consumers of criminal statistics and, as a matter of fact, must lean heavily on them, since they are for certain purposes absolutely essential to our work, we have largely limited ourselves to pointing out their weaknesses but have made no organized efforts to improve them. It is the professional responsibility of the scientist to do everything in his power to make his laboratory material serve his purposes. Mr. Clinard in commenting on the present inadequacy of criminal statistics, says: What is needed for research is the collection of data on meaningful items which will enable us to find answers to some of the theoretical questions about crime. Some of the essential items which are either rarely or never included in statistical enumerations are age at the time and nature of each offense, the number and age of companions in each offense, ... extent of mobility, residence in what local area of THOBSTEN SELLIN [Vol. 41

the city, sociological analysis of degree of criminal development, and psychiatric analysis. This statement is cited merely for illustrative purposes. Criminologists from all disciplines concerned should, in company with professional criminal statisticians familiar with the practical problems of organizing the reporting and processing of criminal statistics, attempt to make criminal statistics more useful as a research source. This would at the same time make such statistics more valuable for the administration of justice which they were primarily developed to serve.

VII. Before we make some brief comments on the techniques of the sociologist concerned with etiological problems, we might glance at the fundamental patterns into which his researches, as well as those in other disciplines fit. These patterns are well illustrated by the sociologist's studies, which are either static (cross-sectional) or dynamic (historical or genetic). In researches of the static type he wants to know: a) if a given trait or characteristic of a group of offenders is more or less common than in a group of non-offenders or in one or more different groups of offenders; or b) if the number of offenses or offenders in a group or area with certain characteristics is greater or less than the number in a group or area having different characteristics or in the same group or area after the characteristics in question have been modified or have disappeared. In researches of the dynamic type he wants to know if changes, i.e., the rise, fall or fluctuations during a period of time in the number of offenses or offenders in a group or area with certain characteristics differ from the corresponding changes in a group or area with different characteristics or if such changes bear any relationship to changes in the contemporary social or other phenomena which are assumed to affect the life of such groups or in such areas. These patterns, of which only the crudest outline has been presented, are but rarely satisfied. For instance they implicitly require the use of comparative analysis, i.e. of control groups. From this point of view much research, not only sociological, is incomplete. Nothing is more common than to find studies of offenders or offenses without any attempt to relate the data obtained to corresponding data on non-offenders or to data which might throw light upon the relationship of the findings of the study to some aspect or historical change of contemporary social life in the locality or during the period covered by the investigation. It 1950] STUDY OF CBIMINOLOGY

is even more rare to find studies that in the use of control groups or data have employed rigorous inethods of securing adequate groups for comparison with the groups or data studied. Both Dr. Kunter and Dr. Clinard stress the need for adequate control groups. Once the sociologist has decided on what he wants to study and what sources he will use he has to adapt his technique of investigation to these sources. If he has decided to study the individual offender, perhaps in a fairly large number, he may use the techniques of the clinician: the interview, through which he gathers directly the information he desires from the offender himself or from those who know him or tests, such as social attitude tests, or other objective tests which he as sociologist may have devised. He may secure autobiographical documents from the offender, and search in various sources of recorded data for information about him. He will also verify the information secured by communi- cating with agencies or persons who can assist him in this task, etc. In some types of studies he may rely on questionnaires, or on rating scales, which should, to be useful at all, be developed with great skill, care and understanding. In other studies he may have great need of data which have already undergone a statistical analysis by others, such as criminal statistics. In all etiological studies he will ultimately, when he reaches the stage of analyzing his data, need a much more than superficial knowledge of statistical principles and statistical analysis. This knowledge is extremely important today even in the planning of research, for we are of necessity compelled to use sampling procedures in most researches. Many researchers that object to sampling appear to ignore that they are in fact using samples even when they study all recorded offenses of any type or convicted offenders of this or that class. A fundamental weakness of much research in this field is due to the failure of taking into sufficient account the fact that they are based on biased (i.e. unrepresentative) samples resulting from the selective processes of criminal justice. The few reports presented are agreed on the need for the improvement of sociological research techniques. All of them also lay stress on the need for more detailed studies of specific types of criminal conduct rather than on criminality or criminals in general. This demand seems so obviously sound that it requires no elaboration.

VIII. It is impossible to provide in a brief expos6 of this type any but the most general suggestions of what sociologists should try to do to improve and expand their researches on the etiology of criminal conduct. Some 422 THORSTEN SELLIN [Vol. 41 basic needs have already been mentioned here and there in this paper: improvement in the communication of research findings, the need for team work with researchers from other disciplines, the need for a super- legal definition of deviant conduct, the need for more useful criminal statistics, etc. Both Mr. Clinard and Mr. Kunter call for more ecological studies and the former urges that more attention be paid to the criminality of what might be called the upper social classes. The development of better sociological testing devices, in close cooperation with the psychologist, is needed. Finally, there is a large field of research opening through so-called "prediction" studies, of which a large number has been made in the United States and some in Germany. Such studies of offenders have in the past been looked upon as useful for an understanding of the defects in penal treatment and for an improved administration of criminal justice, but they obviously have great value for an understanding of criminal conduct. In fact, as the techniques of such studies improve, it will be found that their greatest scientific contribution will be in the field of criminal etiology, even though they may have a more immediate usefulness in the practical field of penal administration. These studies have demonstrated the capital importance of following the history of groups of individual offenders over a long period of time. Their costliness is their only drawback, but we may perhaps hope that some day nearly as much public funds will be devoted to the scientific study of human conduct as is now expended on the improvement of lethal weapons.